claim
Model Variability, as a metric for language model evaluation, captures the variation in hallucination behavior across different models for the same prompt type, representing intrinsic model bias or instability.

Authors

Sources

Referenced by nodes (1)